Research on TRIZ/QFD-Based AIGC-Assisted Process Design

Purpose To address the challenges of low solution convergence efficiency and the imbalance between functionality and aesthetics in Artificial Intelligence Generated Content (AIGC)-driven design processes, this study proposes a TRIZ/QFD collaborative framework for AIGC-aided design. The framework aims to enhance design efficiency and solution quality through structured innovation mechanisms, resolving the core contradiction between user requirements and technical implementations. Methodology First, a human-AI collaborative demand analysis framework was constructed using GPT-4, where user requirements were extracted through focus group interviews and quantified via the entropy weight method. Subsequently, a demand-feature mapping matrix was developed using the QFD House of Quality (HOQ) to identify three critical technical contradictions. The AutoTRIZ tool was then employed to match TRIZ contradiction matrices and generate inventive principle-driven solutions. Finally, these principles were translated into Stable Diffusion-interpretable prompts to produce candidate solutions, which were evaluated through a Pugh matrix and multidimensional entropy-weighted scoring. Conclusion The TRIZ/QFD collaborative framework significantly improves the scientific rigor and user orientation of AIGC-generated solutions by enabling quantitative demand mapping and structured contradiction resolution. This approach reduces designers’ learning curves while providing methodological references and practical paradigms for AI-driven industrial design.